Computer device for detecting an optimal candidate compound and methods thereof

Abstract

The invention relates to a method for a computer device, for detecting an optimal candidate compound based on a plurality of samples comprising a cell line and one or more biomarkers, and a plate map configuration, wherein the plate map configuration is providing locations of samples comprising cell lines exposed to one or more biomarkers and different concentrations of a candidate compound forming at least one concentration gradient, the candidate compound being comprised in a plurality of can-didate compounds, said method comprising generating ( 310 ) phenotypic profiles of each concentration gradient of each of the plurality of candidate compounds at a plurality of successive points in time to form a plurality of compound profiles, wherein generating phenotypic profiles comprises the steps obtaining ( 312 ) image data depicting each sample comprised in the concentration gradient, generating ( 314 ) a class-label and a class for each cell of the samples based on the image data, detecting ( 320 ) the optimal candidate compound by evaluating a comparison criterion on the plurality of compound profiles. Furthermore, the invention also relates to corre-sponding computer device, a computer program, and a computer program product.

The invention claimed is: 
     
       1. A computer-implemented method for generating class-labels and/or classes for cells comprising:
 obtaining image data associated with a plurality of samples of a cell line treated with a biomarker and subjected to experimental perturbation; 
 segmenting, by a processor, cell objects from the obtained image data; 
 generating a set of class-labels corresponding to each of the cell objects; 
 running an unsupervised clustering algorithm on the processor to classify each of the cell objects with a corresponding one of the set of class-labels; 
 displaying at least one cell object and the respective class-label; 
 receiving data indicative of an operation on at least one class-label; and 
 performing the operation on the class-label by the processor based on the received data. 
 
     
     
       2. The method of  claim 1 , wherein experimental perturbation comprises at least one of: inhibiting enzyme activities in cells, promoting enzyme activities in cells, genetic modification of cells, or applying a candidate compound to cells. 
     
     
       3. The method of  claim 2  wherein applying a candidate compound comprises applying different concentrations of the candidate compound to the cells forming a concentration gradient. 
     
     
       4. The method of  claim 1 , wherein the operation comprises one of add class, delete class, split class, and merge class. 
     
     
       5. The method of  claim 1 , wherein displaying at least one cell object and the respective class-label comprises displaying an exemplary cell object and the respective class-label and/or class comprises:
 calculating a confidence measure of how well each cell object represents the class-label; and 
 selecting the exemplary cell object as the cell object with the highest confidence measure. 
 
     
     
       6. The method of  claim 1  wherein the class-label is indicative of cell health status or cell cycle status. 
     
     
       7. The method of  claim 1 , further comprising generating phenotypic profiles of a concentration gradient for a candidate compound at each of a plurality of successive points in time to form a compound profile. 
     
     
       8. The method of  claim 7 , further comprising forming a first collection of cytometric parameter sets by generating a cytometric parameter set for each cell object based on the image data. 
     
     
       9. The method of  claim 1 , further comprising storing at a memory a phenotypic classification model based upon the class-labels, wherein the phenotypic classification model is configured to map a cytometric parameter set to a class, based on a parameter similarity function. 
     
     
       10. The method of  claim 9 , wherein the parameter similarity function is a multi-dimensional correlation function configured to operate over two or more cytometric parameter sets. 
     
     
       11. The method of  claim 10 , wherein the parameter similarity function is a machine learning algorithm selected from the group consisting of: self-organizing maps, auto -encoders, Ward Clustering, K-Means Clustering, and t-SNE Dimensionality Reduction. 
     
     
       12. The method of  claim 1 , further comprising:
 detecting an optimal candidate compound selected from one or more reference compound profiles and based on the experimental perturbation; and 
 calculating a multi-dimensional differential value for each of the one or more reference compound profiles based on the experimental perturbation. 
 
     
     
       13. A system comprising:
 a computer-readable medium storing instructions that, when executed by a processor:
 obtain, by the processor, image data associated with a plurality of samples of a cell line treated with a biomarker and subjected to experimental perturbation; 
 segment cell objects from the obtained image data; 
 generate a set of class-labels corresponding to each of the cell objects; and 
 run an unsupervised clustering algorithm on the processor to classify each of the cell objects with a corresponding one of the set of class-labels; 
 render for display at least one cell object and respective class-label from the obtained image data, wherein the rendered displayed at least one cell objects are segmented based upon the classification from the unsupervised clustering algorithm; 
 receive data indicative of an operation on at least one class-label; and 
 perform the operation on the class-label by the processor based on the received data. 
 
 
     
     
       14. The system of  claim 13 , wherein the operation comprises one of add class, delete class, split class, and merge class. 
     
     
       15. The system of  claim 13 , wherein displaying at least one cell object and the respective class-label comprises displaying an exemplary cell object and the respective class-label and/or class comprises:
 calculating a confidence measure of how well each cell object represents the class-label; and 
 selecting the exemplary cell object as the cell object with the highest confidence measure. 
 
     
     
       16. The system of  claim 13 , wherein the class-label is indicative of cell health status or cell cycle status. 
     
     
       17. The system of  claim 13 , wherein the processor further generates phenotypic profiles of a concentration gradient for a candidate compound at each of a plurality of successive points in time to form a compound profile. 
     
     
       18. The system of  claim 17 , wherein the processor further forms a first collection of cytometric parameter sets by generating a cytometric parameter set for each cell object based on the image data. 
     
     
       19. The system of  claim 13 , further comprising storing at the computer -readable medium a phenotypic classification model based upon the class-labels, wherein the phenotypic classification model is configured to map a cytometric parameter set to a class, based on a parameter similarity function. 
     
     
       20. The system of  claim 13 , wherein the processor is further configured to:
 detect an optimal candidate compound selected from one or more reference compound profiles and based on the experimental perturbation; and 
 calculate a multi-dimensional differential value for each of the one or more reference compound profiles based on the experimental perturbation.